SIGMA

SIGMA or STEAM-SIGMA is a Java-based tool designed to construct accelerator magnet models using the COMSOL Multiphysics® API (https://www.comsol.com/). To enhance its usability and enable seamless integration with Python-based projects, we have developed a repository called pySIGMA.

STEAM-pySIGMA serves as a Python wrapper for SIGMA, offering users the convenience of interacting with the tool using Python. Additionally, it orchestrates the inputs and creation of the generated COMSOL model, further simplifying the process of utilizing SIGMA's functionalities within Python projects.

The integration between SIGMA and Python is made possible through the use of the STEAM-SDK, an essential component responsible for constructing input files required for pySIGMA. The STEAM-SDK generates the necessary data, enabling pySIGMA to interface effectively with SIGMA.

To allow Python to access SIGMA's core functionalities, we package the relevant functionality into a Java Archive file (.jar). This .jar file is placed within the STEAM-pySIGMA repository and utilized through the py4j library.

A visual representation of the SIGMA workflow is shown in the picture below. 1. The shared yaml input file from steam-sdk together with data from ROXIE are used in STEAM-SDK to generate the three relevant input files: .yaml file contains all variables that can be edited within a simulation. .set file contains variables that in first hand is not supposed to be changed. the .geom file contains the geometry from ROXIE which is outputted by ParserRoxie.py in STEAM-SDK. 2. The input files are used in pySIGMA to build geometry, domains and materials (input from material library). 3. pySIGMA outputs java files that contain the COMSOL model for the specific magnet. These can be compiled through JDK (Java Development Kit) and read into COMSOl all within pySIGMA. SIGMA

The pySIGMA documentation below describes the input file parameters.

Getting started:¶

To begin exploring SIGMA and using pySIGMA, follow the steps outlined below:

Ensure you have JDK 8 installed on your system. We currently only support this version:
- JDK 8.281: https://www.oracle.com/java/technologies/javase/javase8u211-later-archive-downloads.html
You must have COMSOL version 5.3a or 6.0 installed on your machine. Please note that other versions are not supported.
Ensure you have access to the STEAM-Material library. If you don't have access, kindly contact the steam-team through the email provided below.
Clone the pySIGMA repository using the link provided below.
Navigate to the file test/setting.SYSTEM.yaml within the cloned pySIGMA repository. Update the following paths accordingly:
- comsolexe_path: Set this to the path of your COMSOL.exe file.
- JAVA_jdk_path: Set this to the path of your JDK installation.
- CFunLibPath: Set this to the path of the steam material library.
Now, run the test script test_T0_QH_Magnet.py to ensure the model builds correctly and try opening it in COMSOL.
Run the test script test_full_magnet.py and verify if MQXA builds correctly.

Once you have confirmed that the pySIGMA tests run successfully on your local setup, you can proceed to STEAM-SDK and use BuilderSIGMA to create input files for any magnet of your choice. You can execute pySIGMA from STEAM-SDK using DriverSIGMA.py (pySIGMA is a dependency in STEAM-SDK).

To further explore and test the tool, you can use the STEAM-Framework and its analysis files. Clone the repository sigma_verification: https://gitlab.cern.ch/steam/analyses/sigma_verification and run the analysis_loop script.

Status of SIGMA¶

SIGMA only work for Rutherford cables and the following sheet shows which simulations are currently working: SIGMA

SIGMA (Java)¶

SIGMA is able to produce and solve COMSOL models of the following types: * Magnetostatic field solver * Transient magneto-thermal coupled simulations (Quench Simulations)

Please observe that the layers must be connected manually in the COMSOL model for Quench Simulations, see the SIGMA manual.

Below you can see examples of generated and solved COMSOL models using SIGMA. SIGMA

SIGMA

Documentation¶

SIGMA documentation

References (for SIGMA and pySIGMA)¶

[1] L. Bortot, B. Auchmann, I. Cortes Garcia, A. M. Fernandez Navarro, M. Maciejewski, M. Prioli, S. Schöps, A. Verweij, “A 2-D Finite-Element Model for Electro-Thermal Transients in Accelerator Magnets”, IEEE Transactions on Magnetics, 2017

[2] STEAM: Simulation of Transient Effects in Accelerator Magnets, https://cern.ch/steam